Mobile devices often include cameras, so they are commonly used to capture images. The cameras of mobile devices are designed for capturing natural images but, with the advancement of high resolution cameras, many mobile devices are used to capture document images as well, thus replacing devices dedicated to document scanning Generally, a mobile device captures a photo of a document, converts the captured photo to a document format, and saves the document in the document format at the device. The mobile device provides a fast and easy way to capture receipts, business cards, business documents, a book cover, whiteboards, and the like.
In contrast to other types of document scanners, a mobile device is best positioned a certain distance away from a target document because of the optical nature of the camera, similar to optical cameras optimized for capturing natural images. Due to this gap between the mobile device and the target document, the captured image is subject to problems specific to the device, such as irregular lighting conditions, dark shadows, and potential glare. For example, lighting conditions may be conditioned on ambient light sources in proximity to the mobile device. Also, dark shadows may be caused by objects positioned between a light source and the target document, such as the mobile device itself. In addition, potential glare may be caused by light reflected from the surface of the target document.
Conventional devices address the problems associated with luminance, shadows, and glare for scanned documents with post-processing techniques, i.e., processing of images after capture. Captured images may be processed by image editing and graphic design applications in an attempt to enhance the images after-the-fact. Unfortunately, the capabilities of these applications to enhance the images are limited to whatever information is available after the images are captured.
Beyond post-processing, some conventional applications for mobile devices attempt to compensate for low ambient light conditions and intrusive shadows for document images. However, these conventional applications fall short for many situations. For example, the light sensors for the mobile devices often fail to identify illumination issues for document capture, because the light sensors are calibrated for capturing natural images. Also, these conventional applications have difficulty with removing dark or hard shadows from document images, thus creating problems with processing the document images tainted with these shadows. Thus, conventional applications still have limitations when attempting to address the problems of capturing document images.